WDS Configuration

A WDS network can be configured in 2 main ways.  The first (and most preferred way for network throughput) is to have all WDS APs (access points) on a wired backbone, allowing all clients to access the WDS AP wirelessly.  The second is to have the main WDS AP wired to the network backbone and to have the remaining WDS APs connect wirelessly to the main WDS AP.  There are of course variants to these main two configurations.  Ultimately, using a PoE switch to connect all WDS APs to the network backbone is preferred as it allows for a single Ethernet cable to provide networking and power to the access point.

Along with the issues of interoperability comes the speed-related issues that WDS has – if a WDS-enabled AP is used to communicate to both another WDS-enabled AP as well as to client machines (such as laptops and PDAs), then the throughput is roughly halved.  For high-bandwidth Wi-Fi networks, therefore, WDS requires a wired backbone for all WDS Access Points.

In addition to the speed issues, security configuration is limited when using WDS.  In a WDS environment you cannot utilize dynamic encryption keys and are therefore limited to WEP or WPA-PSK, neither of which are at the leading edge of Wi-Fi security (although, at this point, WPA-PSK is rather decent).  802.1x over WDS is simply not possible.  There is currently a Draft 802.11s amendment to enhance mesh networking (such as, but definitely not limited to WDS) which will hopefully cater for dynamically assigned and rotated (changed) security keys.